Differences in Measures of Strength and Dynamic Balance among Individuals with Lower-Limb Loss Classified as Functional Level K3 Versus K4

Abstract

Objective:

For individuals with lower-limb loss (LLL), functional mobility (i.e. K-level) classification can be subjective. Performance-based outcome measures (OM) improve the objectivity of K-level assignment; therefore, this study aimed to determine differences in functional strength- and dynamic balance-based OM performance between K3- and K4-classified adults with LLL.

Design:

Prosthetists used subjective information and prosthetic componentry to determine K-levels prior to OM testing for adults with a unilateral transtibial (n=50) or transfemoral amputation (n=17). OM [i.e., 5-Times Sit-to-Stand Test (5xSTS), Figure-of-8 Walk Test (F8WT), 360° Turn Test (360TT), and modified Four-Square Step Test (mFSST)] were administered by a blinded examiner. Univariate analyses of variance (ANOVAs) were used to evaluate between-subgroup differences.

Results:

K4-classified participants with a unilateral transfemoral amputation performed better on all outcome measures when compared with K3-classified peers, whereas K4-classified individuals with a transtibial amputation performed better on the modified Four-Square Step Test compared with K3-classified peers (P < 0.050).

Conclusion:

K4-classified individuals demonstrated greater lower-extremity functional strength and better dynamic balance compared to K3-classified peers. To assist with K-level classification, clinicians should consider selecting OM that objectively differentiate between K-levels (i.e., mFSST for those with a unilateral transtibial or transfemoral amputation; 5xSTS, F8WT, and 360TT for those with a transfemoral amputation).

Introduction

In 2005, an estimated 1.6 million individuals were living with lower-limb loss (LLL) in the United States, of whom 54% had experienced an amputation due to dysvascularity (e.g. diabetes mellitus, peripheral vascular disease).¹ Given the increasing rates of dysvascular disease in the general population, this number is expected to more than double by 2050¹, leading to an increased demand for lower-limb prostheses. Among the growing number of individuals living with LLL, prosthesis use is associated with better functional outcomes, including greater independence and increased self-reported quality-of-life;² however, changes in healthcare requiring detailed documentation of medical necessity have adversely impacted reimbursement rates for initial and replacement prosthetic devices.³ Denials of prosthesis coverage from third-party payers commonly result from a lack of sufficient evaluation and documentation of an individual’s current or anticipated functional mobility level when using a prosthesis, which must match the function of the prescribed prosthesis.^4,5

To identify ambulatory potential and ensure appropriate prescription of initial and replacement prosthetic componentry, Medicare Functional Classification levels (i.e. K-levels) ranging from K0 to K4 are assigned to patients with LLL.^5,6 K0 is the lowest classification level and indicates an individual lacks the ability or potential to safely transfer or ambulate with or without assistance, whereas K4, the highest classification level, indicates an individual demonstrates the ability or potential for prosthetic ambulation that exceeds basic ambulation skills.^5,6 According to the Centers for Medicare and Medicaid Services (CMS), K-level classification is used to help determine medical necessity and justify reimbursement of a range of prosthetic components.⁷ For example, prosthetic joints with microprocessor technology, which aid community ambulation by providing variable resistance to knee motion during mobility tasks,⁸ are considered more advanced and require higher (i.e. K3 or K4) classification and documentation of functional mobility skills exceeding basic ambulation.⁴ K4-level componentry requires particularly extensive documentation of high functional mobility potential, given the large cost associated with componentry designed to accommodate high physical activity demands.⁴ In contrast, prostheses with limited joint motion, constant friction, or manual locking mechanisms are considered more suitable for individuals with lower functional mobility potential (i.e. K1- or K2-classified), as these devices promote overall stability and safety when transferring or ambulating using a prosthesis.⁹

While K-level classification requires comprehensive assessment of functional mobility potential, prosthetic evaluation components have not been standardized.^4,9 To improve prosthesis prescription, CMS has recommended several examination components (e.g. assessment of upper and lower extremity strength, joint range-of-motion, and walking and balance tests) for use when evaluating individuals with LLL.⁷ While many outcome measures (OM) have been shown to be reliable and valid for assessing physical performance (i.e. walking and balance) among adults with LLL,^5,10 interpretation of performance, as it relates to K-level assignment, is still largely subjective.⁹

Differentiating between K3- and K4-level assignments ensures individuals with exceptionally high functional mobility potential receive prosthetic componentry that matches their needs. Recently, differences in walking performance (i.e. scores on the 10 Meter Walk Test and 6 Minute Walk Test) and performance on the Amputee Mobility Predictor, an assessment of functional mobility potential developed for prosthetic evaluation purposes,⁵ were reported between individuals with LLL classified as K3 versus K4.¹³ Specifically, individuals classified as K4 walked faster at both self-selected and fast speeds, walked further during the 6 Minute Walk Test, and scored higher on the Amputee Mobility Predictor.¹³ While these findings suggest outcome measure performance may assist with distinguishing between K-levels, little evidence exists regarding between-group differences in functional strength, dynamic balance, and complex (i.e. non-linear) walking performance among individuals classified as K3 versus K4. By definition, individuals who demonstrate K3- or K4-level ambulatory skills are capable of community ambulation, which requires mobility skills beyond linear, level walking (e.g. turning, obstacle negotiation).^14,15 Thus, exploration of differences in physical performance measures in higher-functioning adults with LLL may assist clinicians with objective K-level assignment for community-ambulators.

Furthermore, evaluation of turning ability is critical during comprehensive prosthetic evaluations, as nearly all mobility tasks performed during community ambulation involve turning.¹⁶ Conflicting evidence regarding the impact of turn direction on turn performance exists for adults with LLL. For example, it has been reported individuals with a transfemoral amputation turn more slowly toward the prosthetic limb compared to the sound limb during a functional mobility task, while individuals with a transtibial amputation have no difference in turning toward versus away from the prosthesis.¹⁷ Further investigation into the impact of turn direction on turning performance is warranted to determine best practice when administering OM that incorporate turns.

When examining functional strength, dynamic balance, and complex walking skills, interpretation of OM data is a critical component in evaluating the patient’s functional mobility, i.e. K-level. The primary aim of this study was to determine whether differences in OM assessing functional lower extremity strength and dynamic balance exist between adults with a unilateral LLL classified as a K3- versus K4- functional level by a prosthetist blinded to OM results. The authors hypothesized individuals classified at the K4-functional level would demonstrate better performance on select functional strength and dynamic balance OM, when compared to those classified as K3. A secondary aim was to assess if there was a difference in turn performance among adults with unilateral LLL when turning toward the prosthetic side as compared to the sound limb side. The authors hypothesized individuals with a unilateral LLL using a prosthesis would demonstrate worse turn performance toward the prosthetic side.

Methods

Study Design and Sample

Participants were recruited from July to August of 2017 through evaluations at the University of Delaware in the Department of Physical Therapy, local prosthetic clinics, and the 2017 Amputee Coalition National Conference in Louisville, Kentucky. Eligibility criteria included age ≥18 years, presence of a unilateral transfemoral or transtibial amputation that occurred ≥1 year prior, and current use of a prosthesis. Participants were excluded if they had a sound limb amputation, were medically unstable (e.g. presence of uncontrolled blood pressure) or had compromised integrity of their residual limb. All participants signed a written informed consent approved by the Human Subjects Institutional Review Board at the University of Delaware prior to data collections. This study conforms to all STROBE guidelines and reports the required information accordingly (see Supplementary Checklist).

Overview of Procedures

Following the informed consent process, participants were asked demographic and amputation-related questions (e.g. amputation etiology, time since initial amputation, years of prosthesis use). Prosthetists from Independence Prosthetics and Orthotics, Inc. determined each participant’s K-level based on subjective information derived from the patient interview, including information about community ambulation, prosthesis use, and ability to negotiate environmental barriers, as well as observation of current prosthetic componentry. Following K-level assignment, height and weight (while wearing the prosthesis) were collected, and physical performance testing was conducted by trained examiners from the Delaware Limb Loss Studies research laboratory, who were blinded to prosthetist-determined K-level assignment.

Outcome Measures

Functional strength and mobility were measured using the 5-Times Sit-to-Stand Test (5xSTS), a measure of lower-extremity strength and transitional mobility¹⁸ with established reliability (ICC=0.89–0.99) and validity in various patient populations^19,20. Following demonstration, participants were asked to start in a seated position in a standard-height chair, then “stand up straight” and “sit down” as quickly as possible 5 times without stopping, while keeping the arms folded across the chest. Timing started when the examiner said, “begin” and ended when the participant reached a standing position on the 5^th repetition. Two trials were completed, and the best (i.e. fastest) time was recorded.

Walking ability and dynamic balance were assessed using the Figure-of-8 Walk Test (F8WT), the 360-degree Turn Test (360TT) toward both the prosthetic (360TT-P) and sound limb sides (360TT-S), and the modified Four Square Step Test (mFSST), all of which have established reliability (ICCs=.94, .95, .97, and .86, respectively) among adults with LLL. The F8WT provides information on complex walking ability and accuracy of walking.²¹ Participants were asked to walk in a figure-of-8 pattern around two cones “as quickly and accurately as possible” with or without an assistive device.^21,22 This test has been shown to be valid among various populations, including older adults and individuals post-stroke.^21,22 Two trials of the F8WT were performed, and the fastest time was recorded.

The 360TT can be used to assess turning ability towards both the prosthetic and sound limbs and has established validity among older adults²³ and adults post-stroke.²⁴ To complete the test, participants were asked to “turn 360 degrees as quickly and safely as possible” towards either the prosthetic (360TT-P) or sound (360TT-S) limb. Two trials were performed in each direction, and the fastest times toward each direction were recorded. Performance in both directions (i.e. towards the prosthetic limb and towards the sound limb) has been shown to be reliable;¹² however, it is unknown if differences exist in 360-degree turn performance based on the direction of the turn. Differences in bilateral turning speed are worth exploring, given prior assessments of 180-degree turning among adults with a transfemoral amputation indicating faster turns towards the sound limb.¹⁷

Finally, the mFSST assesses dynamic balance through multi-directional stepping. Participants were asked to “complete a sequence as fast as possible without touching the [taped] lines on the floor,” which were arranged in a “+” sign. All participants completed a practice trial after examiner demonstration, prior to two timed trials; the fastest timed trial was recorded. For this study, the mFSST was chosen as the need to clear canes for a valid test during the FSST increases the potential for a floor effect, as has been previously shown among adults post-stroke²⁵.

Statistical Analysis

All statistical analyses were performed using IBM SPSS Statistics 25 (SPSS, Inc. Armonk, NY). Participants were divided into groups based on amputation level (i.e., unilateral transtibial and unilateral transfemoral). For each K-level subgroup (i.e., K3 versus K4) within each amputation level, descriptive statistics were determined for participant demographic data. Using a Shapiro-Wilk test, OM data were assessed for normality. Univariate analyses of variance (ANOVA) were used to evaluate differences in OM performance between K3- and K4-classified participants for each amputation level (i.e., between-subgroup differences), with significance defined as p<.050. While differences in demographic and amputation-specific information were evaluated, personal factors such as age and amputation etiology were not considered as covariates when evaluating subgroup differences in OM performance given established associations between increased age, vascular amputation etiology, and decreased physical function.²⁶ Lastly, a sub-analysis was performed using a paired- t-test to evaluate within-individual differences in 360TT performance based on turn direction (toward the prosthetic versus sound limb side) for both K3- and K4-classified participants at the transtibial and transfemoral levels.

Results

Eighty-four participants met the study’s inclusion criteria and agreed to complete the evaluation. During the evaluation, two participants were assigned a K2-level classification and were consequently removed from the dataset. Of those who completed the standardized evaluation, 14 were excluded from the analysis, resulting in a total of 68 participants classified as K3- or K4-level for the between-subgroup analyses (and within-individual turn-performance analysis). During statistical analyses, one participant consistently appeared as an outlier for performance-based OM and was consequently removed, resulting in a final sample of 67 participants (n = 50 with a unilateral transtibial amputation of whom 28 were classified as K3; n = 17 with a unilateral transfemoral amputation, of whom 10 were classified as K3).

Participant demographics are provided in Table 1. For both transtibial and transfemoral groups, no significant differences were found for sex or years of prosthesis use betweenK3- and K4-classified participants. Among those with a unilateral transtibial amputation, participants classified as K3 (as opposed to K4) were significantly older (P = 0.005) and more likely to have undergone an amputation secondary to dysvascularity (P = 0.010). Compared with K4-classified participants with a transfemoral amputation, those classified as K3 were older and had a significantly higher mean body mass index (P = 0.003); however, between-subgroup differences in age did not reach statistical significance (P > 0.050).

Table 1.

Participant Demographics

	Mean (95% CI)		p	Mean (95% CI)		p
	Transtibial			Transfemoral
	K3 (n=28)	K4 (n=22)		K3 (n=10)	K4 (n=7)
Sex, Male*	14 (50%)	13 (59%)	.577	7 (70%)	5 (71%)	.686
Amputation Etiology, Trauma*	16 (57%)	15 (68%)	.010	5 (50%)	3 (38%)	.536
Age, years	50 (46–55)	40 (35–46)	.005	50 (40–61)	38 (26–49)	.090
BMI, kg/m2	30.9 (28.4–33.3)	28.5 (25.8–31.3)	.203	30.0 (27.0–33.0)	23.9 (21.7–26.1)	.003
Prosthesis Use, years**	8.0 (2.0, 19.0)	4.0 (2.0, 14.3)	.275	7.0 (2.0, 11)	10.0 (7.5, 12.0)	.475

^*Data presented as n (% of sample) rather than mean (SD).

^**Data presented as median (25^th percentile, 75^th percentile)

Abbreviations: CI = Confidence Interval; BMI=Body Mass Index.

Between-subgroup differences in physical performance testing are provided by amputation level in Table 2. Individuals with a transtibial amputation classified at the K4-functional level demonstrated significantly faster performance on the mFSST when compared with those classified at the K3-functional level, whereas individuals with a transfemoral amputation demonstrated significantly faster performance on all measures of interest (ie, 5xSTS, F8WT, 360TT-P, 360TT-S, and mFSST) when compared with those classified at the K3-functional level. To determine true differences in OM performance between K3-classified versus K4-classified participants, 95% confidence intervals (CIs), which represent the estimated range in which the true mean score lies, were calculated for all measures. Minimally overlapping 95% CIs for the mFSSTwere observed for both participants with a transtibial and transfemoral amputation, and nonoverlapping or minimally overlapping 95% CIs between K3- and K4-level subgroups were found for all other OM among participants with a transfemoral amputation. No significant differences were found for within-individual analyses between 360TT-P and 360TT-S performance when analyzing all participants in each amputation level group (transtibial: t = −1.587, P = 0.199; transfemoral: t = −.218, P = 0.830) or when subdividing by K-level classification at the transtibial (K3: t = −.587, P = 0.562; K4: t = −1.850, P = 0.079) or transfemoral (K3: t = −.253, P = 0.806; K4: t = .026, P = 0.980) level.

Table 2.

Measures of Strength and Dynamic Balance

	Mean (95% CI)		p	Mean (95% CI)		p
	Transtibial			Transfemoral
	K3 (n=28)	K4 (n=22)		K3 (n=10)	K4 (n=7)
5xSTS, sec	8.85 (8.07–9.63)	8.06 (7.17–8.94)	.182	13.21 (11.90–14.55)	10.15 (8.58–11.72)	.004
F8WT, sec	6.39 (5.94–6.83)	5.80 (5.29–6.30)	.084	8.34 (7.59–9.09)	6.84 (5.94–7.73)	.012
360TT-P, sec	2.35 (2.13–2.57)	2.14 (1.90–2.39)	.212	3.50 (3.13–3.88)	2.62 (2.21–3.04)	.004
360TT-S, sec	2.31 (2.08–2.54)	2.02 (1.76–2.28)	.103	3.47 (3.08–3.85)	2.65 (2.19–3.11)	.009
mFSST, sec	8.82 (8.30–9.34)	7.80 (7.21–8.38)	.012	10.34 (9.47–11.21)	8.67 (7.63–9.72)	.018

Abbreviations: CI = Confidence Interval; 5xSTS = 5-Times Sit-to-Stand Test; F8WT = Figure-of-8 Walk Test; 360TT-S = 360-degree Turn Test - Sound; 360TT-P = 360-degree Turn Test – Prosthesis; mFSST = modified Four Square Step Test

Discussion

While previous studies have assessed functional strength and dynamic balance performance among individuals with LLL compared to individuals with intact limbs²⁷ or between individuals with transtibial versus transfemoral amputations,¹⁷ this is the first study to assess balance performance among participants classified by K-level. In support of the primary hypothesis, significant between-subgroup differences were found for the mFSST among individuals with a transtibial amputation and for all OM of interest (ie, 5xSTS, F8WT, 360TT-P, 360TT-S, and mFSST) among individuals with a transfemoral amputation. Ultimately, these findings suggest K3-classified individuals with a transfemoral amputation have both decreased functional strength and poorer dynamic balance when compared to K4-classified peers, while K3-classified individuals with a transtibial amputation have poorer dynamic balance compared to K4-classified peers. Provided means and 95% confidence intervals for K3- and K4-groups may serve as preliminary references for clinicians using the abovementioned OM, assisting in more objective functional level classification among higher-functioning adults with LLL.

When conducting a clinical evaluation for K-level assignment and subsequent prosthesis prescription, clinicians should consider a battery of psychometrically-sound OM to comprehensively assess various aspects of functional mobility.¹³ Previous research in several patient populations, e.g. older adults and adults post-stroke, supports the use of a collection of OM (as compared to a single measure) for determining functional mobility status.^28,29 Data from this study may allow clinicians to incorporate the 5xSTS, F8WT, 360TT, and mFSST, measures of functional strength and balance, into their clinical examinations of adults with LLL. Given 95% confidence intervals, the mFSST may be particularly helpful when it is necessary to differentiate between K3 and K4 functional ability for those with either a transtibial or transfemoral amputation, while the 5xSTS, F8WT, and 360TT may assist in differentiating K3 versus K4-level mobility for individuals with a transfemoral amputation. Study findings complement previous work, which found significant differences in walking performance [i.e. 10 Meter Walk Test, 6-Minute Walk Test) and functional mobility tests [i.e. the Timed Up and Go (TUG), Amputee Mobility Predictor] between individuals with LLL classified as K3 versus K4.¹³ Collectively, such studies provide clinicians with a collection of OM to comprehensively assess prosthesis-enabled mobility.

Mean score differences in 5xSTS performance for K3 versus K4 subgroups (transtibial: 0.79 seconds; transfemoral: 3.06 seconds) exceed the standard error of measurement of 0.60 seconds reported for adults with Parkinson’s disease¹²; however, differences were only statistically significant for those with a transfemoral amputation, indicating that clinicians should consider the 5xSTS during higher-level prosthetic evaluations to justify transfemoral rather than transtibial componentry. Specifically, based on 95% CIs, scores of 11.72 secs or less may indicate K4-level mobility potential for individuals with a transfemoral amputation. As the 5xSTS requires a combination of lower-extremity strength, trunk control, dynamic balance, and transitional movement speed,³⁰ findings indicate K3-classified individuals with a transfemoral amputation likely have greater deficits in these areas when compared to K4-classified peers. Exercise programs focusing on strengthening of the hips, quadriceps, and trunk, as well as postural control and movement speed (e.g. agilities), may be beneficial for individuals with a transfemoral amputation hoping to improve their functional level classification from K3 to K4.

Similarly, K4-classified participants at both amputation levels demonstrated faster F8WT performance compared to K3-classified peers. While mean score differences between K-levels exceed the SEM of 0.16 seconds¹² (transtibial: 0.60 seconds; transfemoral: 1.60 seconds), Between sub-group differences were only significant among those with a transfemoral amputation; thus, the F8WT may be most appropriate for differentiating K3- versus K4-level assignment when evaluating individuals with a transfemoral amputation. Based on 95% CIs, a preliminary cut-off score for clinical assessment of K3 versus K4-mobility level among patients with a transfemoral amputation is 7.57 seconds (i.e., patients who can complete the F8WT in ≤7.57 seconds demonstrate potential for K4-level mobility). Compared to individuals classified as K3, faster F8WT times reflect improved obstacle negotiation and more efficient complex walking skills among K4-classified individuals with a transfemoral amputation, highlighting the need for more advanced prosthetic equipment to allow faster performance of dynamic movements.

Planned 360TT analyses found no within-individual difference for turning direction regardless of amputation level or K-level classification subgroup at the transtibial or transfemoral level. (2.14–3.50 seconds) and 360TT-S (2.02–3.47 seconds) scores for participants with either a transtibial or transfemoral amputation are largely consistent with previously-published results among adults with LLL (360TT-P: 2.66–2.98 seconds; 360TT-S: 2.52–2.86 seconds).¹² Findings are comparable to those among individuals post-stroke, who have demonstrated equal turning ability toward both the affected and unaffected sides,²⁴ and ultimately indicate turn direction may not be a critical factor in turn performance for unlimited community-ambulators with LLL who are using a prosthesis. Discrepancies between this study’s findings and those indicating poorer turn performance toward the prosthetic limb may be attributed to the inclusion of individuals classified as K2 in the previous study,¹⁷ warranting further investigation into the impact of turn direction on turn performance among individuals with LLL who have lower-level ambulatory skills.

Differences in mean 360TT time were found between K3- and K4-classified participants with a transfemoral amputation. Compared with K4-classified individuals with a transfemoral amputation, findings indicate worse turning performance toward both the sound limb and prosthetic limb among K3-classified individuals, potentially reflecting a more cautious approach to turning. Overall, significant differences in mean 360TT-P and 360TT-S scores between K3- and K4-level participants with a transfemoral amputation indicate that the 360TT may be clinically useful for differentiating between K3- versus K4-level mobility among adults with a transfemoral amputation. Specifically, the ability to turn in either direction in less than approximately 3 seconds may indicate K4-level mobility among those with a transfemoral amputation.

Finally, significant differences in mean mFSST scores between K3- and K4-classified participants at both the transtibial and transfemoral levels indicate the clinical utility of this measure for differentiating between K3- and K4-classifications during prosthetic evaluations. Of the OM included in this study, minimal overlap in the 95% CIs for the mFSST suggest this test may be prioritized during prosthetic evaluations among either adults with a transtibial or transfemoral amputation to assist clinicians with distinguishing between K3- and K4-levels.

Study Limitations

While major strengths of this study are the provision of preliminary scores to guide K3- and K4-level mobility classification and examiner blinding, the sample was limited to individuals with either a transtibial and transfemoral amputation who were generally longer-term prosthesis users; thus, these findings may not be generalizable to a more heterogeneous and/or acutely post-amputation population. Further, this study did not control for between-group differences in age, BMI, and amputation etiology during statistical analyses-factors, which may impact physical performance, but appear to be associated characteristics with K-level assignment.

Conclusion

K4-classified adults with a unilateral transfemoral amputation performed better on all OM of interest compared to K3-classified peers, while K4-classified adults with a unilateral transtibial amputation performed better on mFSST than K3-classified peers. Administration of the 5xSTS, F8WT, 360TT, and mFSST may improve objective assessment of functional strength and dynamic balance among adults with LLL and assist practitioners with distinguishing between higher K-level classifications (i.e., K3- versus K4-levels) during clinical evaluations of functional mobility. For each OM, 95% confidence intervals for K3- and K4-functional mobility levels, for both the transtibial and transfemoral amputation level, are provided and may be used to assist clinical decision-making. While use of a battery of OM is encouraged to comprehensively assess functional mobility potential, the mFSST may be prioritized for individuals with either a transtibial or transfemoral amputation when clinicians are time-limited and desire to distinguish between K3- and K4-functional mobility.